PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
August 22, 20250 citationsOpen Access

Dual-mode 0D/2D Spatial Asymmetry Optoelectronic Device Enabled by in situ Microzone Femtosecond Laser Deposition

View Full Paper
ZLZehua LiGZGuisheng ZouJHJinpeng Huo

Key Points

  • The dual-mode optoelectronic device can function as both a photodetector and neuromorphic vision sensor, enhancing versatility.
  • Under photodetector mode, the device senses optical signals up to 250 Hz, demonstrating high-speed performance.
  • The Microzone Femtosecond Laser Deposition method enables precise fabrication of a 0D/2D heterostructure for effective integration.
  • With 96.20% accuracy in MNIST digit recognition, this work highlights the potential for advanced vision technologies.

Abstract

Abstract To develop artificial intelligence and humanoid robotics, it is crucial to fabricate advanced vision systems with high efficiency and versatility. A key challenge is the functional integration of high-speed photodetectors (PDs) and neuromorphic vision sensors (NVSs) into a single device, as current studies suffer from complex architectures or fabrication processes. Hence, we propose a Microzone Femtosecond Laser Deposition (M-FLD) technique that enables the localized, in situ deposition of zero-dimensional (0D) black phosphorus (BP) nanoparticles onto a two-dimensional (2D) MoS2 channel by ablating a micro-scale solid-state target. By M-FLD and h-BN nanomask, we fabricated a spatial asymmetric 0D/2D heterostructure for highly integrated dual-mode optoelectronic device. By changing the direction of Vds, the device can be converted from PD to NVS. Under the PD mode, the device can sense high-frequency optical signals up to 250 Hz. Under the NVS mode, the device’s optical energy consumption per activity is only 191.2 pJ. Based on the sensing and memory capabilities, the device is simulated for MNIST handwritten digit recognition, achieving an accuracy of up to 96.20%. This work provides a flexible and powerful platform for fabricating complex heterostructures, paving the way for highly integrated and reconfigurable neuromorphic vision systems.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Li et al. (2025) studied this question.

synapsesocial.com/papers/68af55ccad7bf08b1eadbf1dhttps://doi.org/10.21203/rs.3.rs-7201050/v1
Ask AI
Helpful
Bookmark
Share
View Full Paper